A role for bivalent genes in epithelial to mesenchymal transition

Abstract

Epithelial to Mesenchymal Transition (EMT) is an important and complex cellular process in embryonic development, wound healing and tumour progression. EMT is often triggered or facilitated through the action of master EMT transcription factors including ZEB1 and TWIST. It has been proposed that prior to malignant progression a subset of tumour cells undergo an EMT which facilitates the development of key malignant properties. In recent years, a clear link between developmental and cancer associated EMT has triggered an increased interest in the role of developmental EMT genes in a cancer setting. Many key developmental genes have a bivalent or poised promoter signature which changes to active during differentiation; this is believed to elicit a faster response time in comparison to an exclusively repressed promoter. Owing to the relevance of EMT in development and pathologies my thesis aimed to answer the core question of whether bivalent genes are relevant in malignant EMT. To answer this question I undertook four aims: 1. To uncover novel bivalent genes that were activated in an EMT. 2. To characterise the expression and role of ADM2, PLEKHO1 and RASA3 in EMT. 3. To characterise ZEB1 isoform expression during EMT. 4. To identify novel ZEB1 target genes. Aim 1: We utilised a common model of human EMT, whereby human mammary epithelial cells (HMLE) undergo EMT in response to TGFβ to become mesenchymal (mesHMLE). We performed ChIP-seq against histone3 lysine4 tri-methylation (H3K4me3) and histone3 lysine27 tri-methylation (H3K27me3) alongside RNA-seq to identify genes that changed from a bivalent to an active epigenetic signature with concomitant changes to RNA levels. From this data set 429 genes that exhibited this epigenetic change including the well-known EMT factors ZEB1 and TWIST1. From this list four genes that were not previously associated with a bivalent signature were studied in detail. Three of these, ADM2, PLEKHO1 and RASA3, had not previously been associated with EMT but had EMT associated properties, while one, ZEB1 was a well-established master EMT transcription factor. Aim 2: Chromatin immunoprecipitation, ChIP-reChIP was used to confirm the change in epigenetic marks for ADM2, PLEKHO1 and RASA3 promoters alongside a combination of molecular and bioinformatics analyses to determine expression levels in epithelial and mesenchymal cell lines. Cellular migration assays where levels of these genes were manipulated showed that ADM2 and PLEKHO1 have both an individual and a synergistic effect on migration while RASA3 did not affect migration. Aim 3: ZEB1 isoform expression during EMT was analysed and it was determined that there was no significant change in relative expression over this process. Aim 4: ENCODE ZEB1 ChIP-seq was analysed to obtain insights into ZEB1 binding and to identify novel potential targets of importance to EMT. Established ZEB1 target genes such as CDH1 and CRB3 were identified and 26 novel genes with known or potential roles in EMT were chosen for further study. Of these, F11R and INADL were found to be ZEB1 responsive. Direct ZEB1 binding was confirmed through ChIP-qPCR. Interestingly, both of these genes are associated with tight-junctions as is the previously established ZEB1 target CRB3. This strongly implicates ZEB1 in mediating tight-junction regulation. While bivalent genes have not been ignored in the field of EMT they have, so far, been understudied. My work addressed this issue and identified ADM2 and PLEKHO1 as novel EMT associated genes that play an important role in migration. I also established ZEB1, a master regulator of EMT, as a bivalently regulated gene. These contributions help establish bivalently regulated genes as a valuable, underutilised resource for the identification of novel EMT genes.